General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. Users may download and print one copy of any publication from the public portal for the purpose of private study or research. You may not further distribute the material or use it for any profit-making activity or commercial gain You may freely distribute the URL identifying the publication in the public portal If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. Downloaded from orbit.dtu.dk on: Aug 26, 2020 Application of proteomics to investigate barley-Fusarium graminearum interaction Yang, Fen Publication date: 2011 Document Version Publisher's PDF, also known as Version of record Link back to DTU Orbit Citation (APA): Yang, F. (2011). Application of proteomics to investigate barley-Fusarium graminearum interaction. Technical University of Denmark.
130
Embed
Application of proteomics to investigate barley … D thesis_Fen Yang.pdfApplication of proteomics to investigate barley-Fusarium graminearum interaction Fen Yang Ph.D. Thesis March
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
Users may download and print one copy of any publication from the public portal for the purpose of private study or research.
You may not further distribute the material or use it for any profit-making activity or commercial gain
You may freely distribute the URL identifying the publication in the public portal If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from orbit.dtu.dk on: Aug 26, 2020
Application of proteomics to investigate barley-Fusarium graminearum interaction
Yang, Fen
Publication date:2011
Document VersionPublisher's PDF, also known as Version of record
Link back to DTU Orbit
Citation (APA):Yang, F. (2011). Application of proteomics to investigate barley-Fusarium graminearum interaction. TechnicalUniversity of Denmark.
1.1 Fusarium head blight in barley and wheat ............................................................................................................. 1
1.3.3 Analysis of post-translational modifications (PTMs) ..................................................................................... 12
1.3.4 Analysis of protein-protein interaction ........................................................................................................... 13
1.3.5 Limitations and expectations in proteomics ................................................................................................... 13
1.4 Application of proteomics in plant-pathogen interactions ................................................................................... 14
1.5 Objectives of the project ...................................................................................................................................... 18
Gene expression of Fusarium graminearum during infection in wheat and barley spikelets
Concluding remarks and perspectives .................................................................................................................... 91
Appendix I ................................................................................................................................................................. 93
Implications of high-temperature events and water deficits on protein profiles in wheat (Triticum aestivum L. cv.
Vinjett) grain
Appendix II ............................................................................................................................................................. 116
Secretome analysis of YAP- or AP-knockout mutant of Fusarium graminearum
Chapter 1 Introduction
1
Chapter 1
Introduction
1.1 Fusarium head blight in barley and wheat
Fusarium head blight (FHB) is a devastating disease of cereals including barley and wheat in humid and semi-
humid climates worldwide (Walter et al., 2009). FHB was first described in 1884 in England and has increased
in Asia, Canada, Europe and South America since then (Stack, 1999). The disease has reached epidemic levels in
several years and causes significant losses of grain yield and quality to be millions of dollars per annum in the
USA alone (Nganje et al., 2004). FHB has been identified by International Maize and Wheat Improvement
Center as a major factor limiting wheat production in many parts of world (Stack, 1999). The more important is
the contamination of the mycotoxins produced by the fungus, which adversely affect grain quality and is an
enormous challenge for cereal breeders and the food or milling industry.
Several Fusarium species including F. sporotrichioides, F. culmorum, F. avenaceum, F. poae and the most
prevalent F. graminearum (teleomorph: Gibberella zeae (Schwein) Petch) are the causal agents of FHB
worldwide (Parry et al., 1995). As a result, F. graminearum quickly become one of the most intensively studied
fungal plant pathogen. The genome sequence released by Broad Institute in 2003 has greatly stimulated the
research activity on F. graminearum. The fungus has a genome size (36.1Mb) and contains genes encoding
13937 predicted proteins distributed over four chromosomes with few repeat sequences (Cuomo et al., 2007;
Trail, 2009). There are 2001 genes not similar to those of any other sequenced organisms and 5812 genes having
homology to genes encoding proteins of unknown function (Trail, 2009). The availability of whole genome
sequence provides the opportunity to study transcription and proteome profiling for identifying essential
elements in pathogenesis and lead to development of new targets for fungal control.
Fusarium graminearum colonizes living host tissue at specific stages and establishes itself in senescent
tissue and debris as saprophytic mycelia due to a brief biotrophic relationship with its host before switching to
the necrotrophic phase (Bai and Shaner, 1994; Goswami and Kistler, 2004). The necrotrophic stage is associated
with an increase in vigour of colonization and eventually plant death leads to thorough colonization of the host
substrate (Goswami and Kistler, 2004). The warm moist weather is favourable for the development and
maturation of conidia and perithecia on the crop debris which produce ascospores (Goswami and Kistler, 2004).
The rain and wind spread the soil-born inoculums, mainly ascospores to the plants (Bai and Shaner, 1994). The
Chapter 1 Introduction
2
abundance of the primary inoculums and weather conditions, mainly moisture and temperature, during and after
anthesis determine the severity of FHB (Bai and Shaner, 1994).
The spikelets of barley and wheat are most susceptible at anthesis or during the early dough stage of grain
development. The preferred infection site of Fusarium graminearum is the tip of the kernel. Fig. 1.1 shows
Fusarium graminearum-infected barley and wheat. However, the development of FHB in wheat and barley is
different. In wheat, the fungal hyphae develop on the exterior surfaces of florets and glumes, possible leading to
direct penetration of the epidermal cell (Bushnell et al., 2003). Alternately, the fungus directly enters the stomata
and underlying parenchyma, exposed anthers and openings between the lemma and palea of the spikelet
(Bushnell et al., 2003). Spread of the fungus among florets is through the vascular bundles in the rachis and
rachilla (Ribichich et al., 2000). It is mainly type I resistance to initial infection rather than type II resistance to
spread of infection within a spike in wheat. By contrast, in barley, the fungus penetrates directly stomata and
grows from abaxial to adaxial side of floral bracts. The internal spread through the rachis is more limited,
indicating the type II resistance (Bushnell et al., 2003). It was reported that infection-related structure
development and other morphological changes were observed around 12 to 24 h earlier in wheat than in barley
(Boddu et al., 2006). The fungus can invade the aleurone layer and grow into the starchy endosperm. The light
and electron microscopy studies have shown during infection of the endosperm the cell walls were macerated,
the protein matrix disappeared and the starch granule structures were changed (Nightingale et al., 1999).
Fig. 1.1. Disease symptoms on F. graminearum-infected spikelets of barley (left, photographed by Jens Due Jensen, University of
Copenhagen) and wheat (right, Goswami and Kistler, 2004). The fourth barley spikelet from the bottom up to six spikelets shows
premature necrosis and brown/grey discoloration. The third wheat spikelet from the bottom shows a darkened necrotic lesion whereas the
second and fifth spikelets demonstrate tissue beaching (Goswami and Kistler, 2004).
Chapter 1 Introduction
3
During the penetration of cell wall and access to the plant nutrient for the growth, F. graminearum can
produce an arsenal of hydrolases such as lipases, xylanases, pectinases, cellulases and proteases as well as the
secondary metabolites mycotoxins such as the major trichothecene and zearalenone which are toxic to human or
livestock consumption (Kang and Buchenauer, 2000; Maier et al., 2006). These secreted enzymes are known to
have an important role in pathogenicity of F. graminearum (Kang and Buchenauer, 2000). The toxin
trichothecene is involved in blocking peptidyl transferase activity at the 60S ribosomal subunit in eukaryotes.
Blocking ribosomal activity can inhibit nucleic acid synthesis and mitochondrial function and interfere initiation,
elongation or termination of protein synthesis as well as have negative effect on cell division and membrane
integrity (Khachatourians, 1990). The most prevalent trichothecene derivatives deoxynivalenol (DON) levels are
regulated in food supplies of many countries. For example, the European Community limits DON levels to 0.5
µg g-1 for cereals and the United States limits DON levels to 1 µg g-1 for finished products for human
consumption (Council for Agricultural Science and Technology, 2003) (Trail, 2009). Although zearalenone
which causes estrogenic effects in animals and humans is of concern to the U.S. Food and Drug Administration,
there are currently no regulatory standards limiting its levels in grain (Council for Agricultural Science and
Technology, 2003) (Trail, 2009). DON is the only mycotoxin shown to be a virulence factor (Trail, 2009) and
mediates partially the shift from the biotrophic to the necrotrophic stage of the fungus (Bushnell et al., 2003).
The accumulation of DON depends on the complicated interactions between the host and fungal genotypes as
well as environmental conditions (Mesterhazy et al., 1999). Generally speaking, there is a correlation between
FHB severity and DON concentration in the infected grain (Bushnell et al., 2003). F. graminearum expresses
genes for DON biosynthesis immediately following the infection of wheat (Jansen et al., 2005). DON causing
tissue necrosis allows F. graminearum to spread into the rachis from florets but is not necessary for the initial
infection in wheat (Jansen et al., 2005; Bluhm et al., 2007). It has been reported that trichothecene non-
producing F. graminearum strain was pathogenic but produced a reduced incidence and severity of infection,
less bleaching heads and less spread in spikelets in wheat comparing to the trichothecene-producing strain
(Proctor et al., 1995; Maier et al., 2006). However, no significant difference in virulence was observed in barley
between these F. graminearum strains (Maier et al., 2006). Spread of the disease is limited and virulence does
not appear to be due to the presence of the toxin in barley (Jansen et al., 2005).
The main approaches for controlling FHB are management practices, fungicide application and
development of resistant varieties. Management practices include crop rotation and tillage practice which can
reduce the fungal survival on the residue and staggering planting of small grain crops (Stack, 1999). FHB
severity can be reduced 50 to 60 % by application of the fungicides at early flowering stage in wheat and at the
early heading stage in barley (Stack, 1999). However, results with application of fungicides can be variable due
to the environmental effects and relationship between the cost and return is the limit factor (McMullen et al.,
Chapter 1 Introduction
4
1997). Using genetic approach to control FHB is the most desirable option. However, since the resistance in
wheat and barley to FHB is a complex and quantitative trait involving the interactions among pathogen,
trichothecenes, environments and genotypes, the development of resistant cultivars is very challenging. Current
breeding strategies focus on combination of desired agronomic traits and type I and type II resistance in addition
to selection for low DON in the kernels (Bai and Shaner, 2004). Quantitative trait loci (QTL) mapping, the
statistical study of the alleles occurring at a locus and the phenotypes that they produce using molecular markers
is a powerful tool to select resistant plants in the breeding program. So far most of the FHB resistant QTLs have
been mapped to the same locations as those associated with morphological traits such as heading date, plant
height, lateral floret size, spike angle and kernel plumpness (Bai and Shaner, 2004). QTL analysis is often
associated with global gene expression profiling to identify the key gene markers involved in plant defense
against infection, providing insights into defense mechanism (Bai and Shaner, 2004). Furthermore, as growth
environmental factors such as temperature, humidity and fertilization can affect the disease severity, adjusting
time and places of planting or changing the amount and type of fertilizers can be the option to control the disease
(Yang et al., 2010a). Application of biological control such as microorganisms Bacillus spp., yeasts and
Trichoderma harzianum is an additional strategy in management of FHB in cereals (Corio da Luz et al., 2003).
Alternatively, transgenic expression of genes for antifungal proteins, genes involved in defense reaction and
genes involved in reduction of DON in wheat and barley is an approach against Fusarium infection (Dahleen et
al., 2001).
1.2 Molecular plant-pathogen interactions
Plant-pathogen interactions have been studies for several years in order to understand how plants and pathogens
recognize each other and differentiate to establish either a successful or an unsuccessful relationship (Mehta et
al., 2008). The pathogens may use the following strategies to attack and colonize the hosts: they produce
hydrolytic enzymes to degrade the cell wall and break down the protein for nutrients, synthesize molecules that
can induce the production of enzyme that degrades cell walls, starch and protein and produce some secondary
metabolites like mycotoxins to interfere host metabolism (Bluhm et al., 2007). Unlike many phytopathogenic
fungi, F. graminearum does not produce specialized infection structures such as appressoria or haustoria which
are invaginated into the host cell plasma membrane forming an intimate interface during interaction (Jones and
Dongl, 2006; Bluhm et al., 2007). Instead, colonization of tissues is facilitated primarily by the production of
cell-wall-degrading enzymes (CWDEs) such as cellulases, pectinases and xylanases as well as proteases (Bluhm
et al., 2007).
Chapter 1 Introduction
5
In plant, the response of extracellular signals form pathogen must be rapid, reliable and specific. The
pathogen infection can initiate very complex chains of reactions in plants that lead to various defense responses.
The recognition of CWDEs or molecules from pathogens which is called microbial /pathogen-associated
molecular patterns (PAMPs) can trigger the plant basal defense including thickening plant cell wall, papilla
deposition, transduction of signals such as phytohormones salicylic acid, jasmonates and ethylene to the other
parts of the plants and synthesis of antimicrobial compounds like phytoalexins and biosynthesis of pathogenesis-
related (PR) proteins and defense-related proteins in both compatible and incompatible interactions (Mehta et al.,
2008; Pieterse et al., 2009). In the incompatible interaction superimposed on the basal defense plants can express
disease resistance (R) proteins recognizing the virulence effectors and induce hypersensitive response (HR),
which is the second response or gene to gene specific resistance to prevent pathogen invasion and disease
development (Jones and Dongl, 2006; Mehta et al., 2008) (Fig. 1.2). A series of biochemical perturbations such
as ion flues, lipid hyperperoxidation, protein phosphorylation, nitric oxide generation, a burst of reactive oxygen
species (ROS) and biosynthesis of antimicrobial compounds are stimulated in HR which keep the pathogen
isolated from the rest of the plant and prevent further damage (Mehta et al., 2008; Pieterse et al., 2009).
Moreover, pathogen may also express some proteins such as superoxide dismutase and catalases to overcome the
plant defense or to inactivate ROS for protecting themselves. Therefore, the interaction between host plant and
pathogen is in a complicated and dynamic manner.
Chapter 1 Introduction
6
Fig. 1.2. Overview of plant–pathogen interactions. Plants process receptors that can activate basal resistance, mediated by pathogen-
associated molecular patterns (PAMPs) or cell-wall-degrading enzymes (CWDEs), which may result in a compatible or incompatible
interaction. In both interactions, several defense-related and biotic stress-responsive proteins are induced. Suppression of plant defenses
by pathogen effectors leads to susceptibility in host plants. Some host plants express resistance (R) proteins, which guard against this
interference and trigger a specific resistance, referred to as the hypersensitive response (HR) (Mehta et al., 2008).
Given the essential role in plant defense, PR proteins have been studied for several years from sequence to
the biological function properties. PR proteins are usually defined as host-specific proteins that are induced in
several plant species during pathological or related situations such as pathogen attack, wounding and abiotic
stress (van Loon et al., 2006). However, it does not state clearly that they have functional roles in defense (van
Loon et al., 2006). PR proteins are often low molecular weight proteins (10 to 40 kDa) which can survive and
remain soluble in harsh environments such as extreme pH and be resistant to proteolytic cleavage due to their
biochemical properties. There are 17 PR families based on amino acid sequences, serological relationship and
biological activity (Table 1) (van Loon et al., 2006). The families were originally identified from tobacco and
also other plant species including barley, wheat, rice and maize and are numbered by the order in which they
were discovered. In each family there can be several different isoforms.
Chapter 1 Introduction
7
.
The specific functions of PR proteins are not fully understood. Various PR proteins have potential
antimicrobial activity and are involved in defense mechanisms against pathogens (van Loon et al., 2006).
Chitinase and β-1,3-glucanase have functions involved in the hydrolysis of fungal cell walls. Peroxidase is an
antioxidant and can function in plant cell wall rigidification. Oxalate oxidase is involved in signal transduction
(Christensen et al., 2002). PR proteins can also be detected during plant development and senescence (van Loon
et al., 2006), which may indicate a more physically protective role of the cellular structures in order to stabilize
sensitive membranes or macromolecules (van Loon and van Strien, 1999). Genetic engineering of plants for
introduction of PR genes by transformation or manipulation of the signals that trigger the expression of PR
proteins may be the approaches to improve plant resistance against pathogen infection. Transgenic wheat and
barley expressing genes encoding chitinase, α-thionin, thaumatin-like protein or β-1,3-glucanase have shown the
enhanced resistance against Fusarium graminearum (Dahleen et al., 2001; Mackintosh et al., 2007; Shin et al.,
2008).
Chapter 1 Introduction
8
1.3 Technologies in proteomics
Genome only represents the first step in the complexity of understanding biological function. Transcripts can not
give complete information on cellular regulations as gene expression is regulated post-transcriptionally and
proteins which are responsible for the cell biological functions are expressed in a highly dynamic and interacted
manner (Dhingra et al., 2005). Thus, it is necessary to determine the protein levels directly. Proteomics is the
systematic study of all the proteins expressed by a genome or by a cell or tissue, particularly their interactions,
modification, localization and functions (Coiras et al., 2008). Currently, proteomics has established itself as an
indispensable technology to interpret the information from genomics and has been most successfully applied in
protein sequencing, protein quantification, post translational modifications (PTMs) and protein interactions
(Aebersold and Mann, 2003).
1.3.1 Proteomics workflow-protein preparation, separation and identification
Proteomics workflow mainly consists of protein preparation, protein separation and protein identification by
mass spectrometry (MS). Protein preparation includes tissue and cell homogenization, protein solubilisation and
denaturation with use of detergents such as 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate,
sodium dodecyl sulphate (SDS), urea and thiourea and removal of impurity in samples such as carbohydrates,
lipids, salts, nucleic acids, etc which can interfere protein separation process. It is impossible to obtain the entire
proteome at once since cellular protein populations have enormous diversity due to function, sequence, physical
properties and relative abundance (Hurkman and Tanaka, 2007a). Proteins extracted strongly depend on the
extraction protocols. The extraction of protein from plant samples is in particular challenging, because plants
cells generally contain i) low amounts of proteins protected by cell walls that require extreme measures to
disrupt, ii) proteases that remain active in the extraction buffer, reducing and altering protein populations and iii)
various non-protein components such as cell wall, storage polysaccharides, lipids, phenolics, salts, nucleic acids
and a broad array of secondary metabolites, which cause streaking and smearing of 2-DE patterns in the
following separation procedure (Hurkman and Tanaka, 2007b). Therefore, the extraction methods should
minimize protein degradation and eliminate non-protein components, which are the key steps for proteomic
analysis, especially when 2-DE is used. In our study cases, the majority of barley seeds proteome is non-
metabolic storage protein hordein which is alcohol-soluble. In order to access to metabolic albumins, protein was
extracted at low salt and neutral pH followed by acetone precipitation for concentration and cleanup of samples
(Østergaard et al., 2004)
Chapter 1 Introduction
9
The traditional approach for separating proteins or peptides is two-dimensional gel electrophoresis (2-DE)
separating hundreds of proteins according to their pI and mass (gel-based approach). Although 2-DE technique is
quite clear and reproducible to show a full picture of protein pattern, it has some limitations such as
solubilisation of membrane proteins, invisibility of very low abundant proteins and segregation of extreme pI
and mass (Ong and Mann, 2005). Poor separation of basic proteins due to streaking of spots is also a limiting
factor in 2-DE (Bae et al., 2003). Currently the most powerful strategy is mono- or multi-dimensional liquid
chromatography (LC) which allows high throughput separation of complex protein or peptides mixtures (gel-free
approach) (Coiras et al., 2008). LC separates proteins and peptides according to their affinity for a stationary
phase when a mobile phase is forced through a fine capillary.
Mass spectrometry consists of an ion source, a mass analyser that measures the mass to charge ratio (m/z) of
the ionized analytes and a detector that registers the number of ions at each m/z value (Aebersold and Mann,
2003). The most common techniques for ionizing samples are matrix assisted laser desorption/ionization
(MALDI) and electrospray ionization (ESI). The mass analysers include time-of-flight (TOF), ion trap,
quadruple and fourier transform ion cyclotron (Aebersold and Mann, 2003). They can be used alone, called MS,
or put together in tandem, namely MS/MS. In single mass analyser peptide ions are generated. In MS/MS
specific precursor ions produced in the initial mass analyser are chosen and fragmentized through collision
resulting in fragment ion spectra.
In general, 2-DE is often followed by MALDI-TOF or TOF-TOF mass spectrometry for analysis of
relatively simple protein samples which are crystallized with matrix before ionization via laser pulses in MS.
This instrument has high sensitivity, resolution and mass accuracy (Aebersold and Mann, 2003). LC is often
coupled to ESI-tandem mass spectrometry (MS/MS) for analysis of samples in solution. This method can detect
low abundant proteins and analyse proteolysed peptides without fractionation of protein samples comparing to 2-
DE-based approach. The strategy in MALDI MS for protein identification is peptide mass fingerprinting which
requires purified protein samples. A list of peptide masses of protein generated by MS is searched against a
database which supplies theoretical peptide masses of proteins. Additionally, the fragment ion masses from
selected precursors can be used to match against protein sequence database when MS/MS instrument is used.
The identification is based on matching score calculated by software algorithms. The MS/MS data enables de
novo sequencing and PTM analysis (Coiras et al., 2008).
1.3.2 Quantitative proteomics
In the term of relative quantification of protein, there are two major approaches. The first one is gel-based
quantitative studies relying on protein labelling for later image comparison (Fig.1.3). The second one is gel-free
Chapter 1 Introduction
10
techniques making use of isotopic or isobaric labelling of proteins or peptides for LC-MS/MS analysis (Gstaiger
and Aebersold, 2009) (Fig.1.3).
Fig. 1.3. Overview of quantitative proteomics workflow. Gel-based approach is based on image analysis of protein spot intensity. Gel-
free approaches such as ICAT, SILAC and iTRAQ are based on isotope or isobaric labelling. The mass shift among the labelled peptides
form different samples will be revealed in MS or MS/MS which allows identification and quantification of proteins.
Briefly, in gel-based approach, proteins are separated by 2-DE and stained by dye. Besides Coomassie
Brilliant Blue staining, there are several fluorescent staining methods that have been developed for the
visualization of 2-DE patterns and detection of less abundant proteins, including sypro ruby staining, silver
staining and Cy-dyes. Sypro ruby staining has a comparable sensitivity with silver staining and allows much
higher reproducibility, wider dynamic range, less false-positive staining (Berggren et al., 2000). Alternately,
protein samples are labelled with different fluorescence dyes (Cy2, Cy3 or Cy5) reacting with lysine in the
protein and separated in one 2-DE gel, which will be fluorescent according to the wavelengths. This technique
Chapter 1 Introduction
11
called differential gel electrophoresis (DIGE) is quite sensitive to reveal the protein amounts, has dramatically
improved the reproducibility and accuracy of quantification, and avoids the gel-to-gel variation in the traditional
2-DE which analyses multiple samples in one gel (Unlu et al., 1997; Coiras et al., 2008). The 2-D gels are
scanned and gel images will be imported into gel analysis software for spot detection, pattern matching, spot
quantification and statistic analysis. The spots significantly changing in intensity among samples will be shown
in this software automatically.
Given the requirement of good resolution of 2-D patterns and poor correlation between spots intensity and
protein abundance in gel-based approach, MS-based quantification techniques by introduction of stable isotope
labelling to peptides have emerged. The isotope-code affinity tag (ICAT) is one of the most employed chemical
isotope labelling methods. Each ICAT reagent consists of a thiol reactive group that reacts with cysteine residues,
an isotope-coded light or heavy linker and a biotin segment for affinity purification. In this system, two samples
are labelled with light and heavy reagents on cycteine thiols, respectively. Then both samples are mixed,
digested by trypsin and separated through chromatography. The relative level of protein in the two samples is
determined by the ratios of signal intensities of the isotopically labelled peptide pairs revealed in MS analysis
(Sethuraman et al., 2004). Different software programs such as proICAT are developed to analyse ICAT labelled
MS data. ICAT is very useful to detect low abundant proteins but has major limitation in only selecting protein
of high cycteine content and low sensitivity to acidic proteins (Gygi and Aebersold, 2000). Another strategy for
determining the differentially expressed proteins from different cellular populations is the stable isotope
labelling with amino acids in cell culture (SILAC) (Ong and Mann, 2005). The cells from different biological
conditions are cultured with media supplemented with isotopically labelled amino acids. After metabolic
incorporation of isotopes during protein synthesis, proteins from each sample are isolated, mixed, digested and
analysed by LC-MS. The ratio of speak intensities of the isotopically labelled peptides in the mass spectrum
reflects the ratio of the protein abundance. This technique is in vivo coding without chemical manipulation and
allows comparison of expression levels of tissue proteome at different physiological states (Gruhler et al., 2005).
Recently, a new approach called isobaric tags for relative and absolute quantitation (iTRAQ) has developed
(Ross et al., 2004). The technique is based on chemically tagging the N-terminus of the digested peptides. The
labelled samples are combined, fractionated by LC and analysed by MS/MS. Fragmentation of the tags attached
to the peptides generates the isotope-encoded reporter ions which provide relative quantitative information on
protein. This technology offers several advantages including the ability to analyse multiplex samples and
increased analytical precision and accuracy, but it need a powerful multidimensional fractionation method for
peptides before MS/MS (Aggarwal et al., 2006). Another method is 18O stable isotope labelling where proteins
are isotopically tagged by means of enzyme-catalysed incorporation of 18O from 18O water during proteolysis
(Aebersold and Mann, 2003). Each peptide generated by the enzymatic reaction carried out in heavy water is
Chapter 1 Introduction
12
labelled at the carboxyl terminal. The labelled peptides from each sample are combined and analysed by MS.
The resulting mass shift between differentially labelled peptide ions permits identification and quantitation of
protein. Furthermore, there is label-free quantification strategy present which is based on spectral counting or
peptide precursors ion intensities obtained in the first MS in a tandem MS. Spectral counting is based on the
assumption that the rate at which a peptide precursor ion is selected for fragmentation in a mass spectrometer is
correlated to its abundance. The spectral counts are then averaged into a protein abundance index for relative
protein quantification. This method works for large and abundant proteins, the number of peptides from small
protein and low abundant proteins is often insufficient for accurate quantification (Gstaiger and Aebersold,
2009).
1.3.3 Analysis of PTMs
Covalent modification of protein such as phosphorylation, glycosylation, acetylation and ubiquitin plays
important roles in the control of the activity, localization and stability of proteins and their interactions with
other macromolecules (Gstaiger and Aebersold, 2009). Functional genomics can not provide experimental
evidence for protein modifications other than protein sequence information for the in silico prediction of
candidate sites of modification (Gstaiger and Aebersold, 2009). MS is shown to be a useful tool for global and
targeted analysis of PTMs of proteins. The principle is that the addition of a chemical moiety to an amino acid
will lead to the mass shift of that residue, which will be revealed in MS. This will allow assignment of
modifications to the peptide in MS or amino acid in MS/MS (Jensen, 2006). Computer programmes are
continuously improved for the systematic scanning and annotation of PTMs from MS and MS/MS data.
However, identification of PTM is quite challenging since the peptides bearing a particular modification can be a
small fraction of the total amount of peptides in the samples (Dhingra et al., 2005; Gstaiger and Aebersold,
2009). Therefore, modification-specific enrichment is usually integrated prior to MS and MS/MS analysis
(Dhingra et al., 2005). Some PTMs can be enriched by derivatization of protein modifications to make them
accessible to chemical solid-phase capture techniques, and other PTMs can be purified using affinity
chromatography or antibodies that are specific for a given modification (Gstaiger and Aebersold, 2009). For
instance, phosphopeptides can be enriched by immunoprecipitation and TiO2 column or immobilized metal-
affinity chromatography which is using Fe (III) or Ga (III) for affinity of the phosphate moiety (Jensen, 2006).
Hydrophilic-interaction liquid chromatography and lectin-mediated affinity are useful methods for purification
of glycopeptides (Jensen, 2006).
Chapter 1 Introduction
13
1.3.4 Analysis of protein-protein interaction
Almost all the proteins function in the context of specific interactions with other proteins. In MS-based protein
interaction experiments, there are three essential components including bait presentation, affinity purification of
the complex and analysis of the bound proteins by LC-MS/MS (Aebersold and Mann, 2003). The workflow is
that protein complexes are purified either using antibodies that recognize the endogenously expressed protein or
using an affinity tag that is fused to the protein of interest, digested and analysed by LC-MS/MS. Compared with
yeast two-hybrid and protein chip-based approaches, this strategy has the advantages that interactions take place
in the native environment and cellular location, and that multicomponent complexes can be isolated and analysed
in a single operation. However, this method can only detect a subset of protein interactions that actually occur
since the biological interactions are of low affinity, transient and dependent on the specific cellular environment
(Aebersold and Mann, 2003).
1.3.5 Limitations and expectations in proteomics
The proteomics research today is severely hampered by the lack of publicly available sequence information for
the not completely sequenced organisms since protein identification is based on availability of the gene or
protein sequences in the public databases (Grossmann et al., 2007). Peptide mass fingerprint is not well suited
for protein identification for these organisms. In order to circumvent this limitation, two different approaches can
lead to an increase in protein identifications. The first one is that MS or MS/MS data is searched against
Expressed Sequence Tags (EST) database or a protein database of an evolutionarily closed related organism,
although EST databases can have only partial coding sequences for the gene and many genes may not be
represented in these databases due to the choice of tissue used for library construction or low mRNA abundance
(Quirino et al., 2010). The second one is amino acid sequence of a peptide extracted from the MS/MS spectrum
for de novo sequencing with aid of software tools. Peptide de novo sequencing can be combined with BLAST
searches to identify peptides on the basis of their homology to peptides in the database (Grossmann et al., 2007).
Other major bottlenecks in proteomics are related to analysis of huge amounts of MS data. Accurate, consistent
and transparent data processing and analysis are integral and critical parts of proteomics workflows (Domon and
Aebersold, 2006). Therefore, advanced bioinformatics tools are required for protein identification and validation
and data repositories, which are challenging and ongoing tasks in proteomics.
Chapter 1 Introduction
14
1.4 Application of proteomics in plant-pathogen interaction
The proteomics techniques are mainly utilised in the following aspects for plant-pathogen interactions including
plant-virus, bacteria, fungi and nematodes interactions: a) detection of plant pathogens; b) comparison of
proteomes and detection of differential protein expression at quantitative and qualitative terms in both plant and
pathogen; c) analysis of PTMs like phosphorylation or modification of proteins induced by the infection. In this
thesis, we will mainly focus on and give the examples of researches on the plant-fungus interactions.
Traditionally, detection of plant pathogens may involve the use of time-consuming cultivation with a
subsequent morphological or biochemical analysis of growth or biochemical characterization, bioassays,
isolation and microscopy (Lopez et al., 2003). However, these techniques cause problems when different
organisms produce similar symptoms in hosts or exhibit similar morphology. Lately, enzyme-linked
immunosorbent assays, polymerase chain reaction, DNA sequencing, fluorescence in situ hybridization and
DNA microarrays are the main techniques for phytopathogen detection (Lopez et al., 2003; Padliya and Cooper,
2006). There are limitations to molecular biology-based or antibody-based techniques as many of these protocols
require reagents that are highly specific for individual pathogens (Kav et al., 2007). The recent proteomics-based
technology has been shown to be able to detect or identify phytopathogens accurately and efficiently which do
not require pathogen-specific reagents (Kav et al., 2007). The technique is based on the identification of
phytopathogen protein in MS analysis which typically is using various publicly available protein databases with
information on phytopathogenic organisms including virus, fungi, oomycete and bacteria (Padliya and Cooper,
2006). Thus, a relative paucity of data in genomic or protein databases pertaining to many pathogens is a
significant obstacle in MS-based studies, but it can be solved partially by cross-species identification (Padliya
and Cooper, 2006). With regard to pathogenic fungi, there are three other complications with identification by
MS comparing to virus and bacteria (Padliya and Cooper, 2006). First, unpredictable PTM can confound the
protein identifications. Second, pathogenic fungus can encode several thousands of proteins but no single protein
will accumulate to high levels for good resolution in MS, especially with a complex plant protein background.
Third, fungi can have several life cycle stages, indicating that the presence of some proteins used in
identification can be in flux. So far, MS-based techniques have been successfully applied in the detection of
phytopathogenic fungi such as Ustilago maydis, Trichoderma harzianum, Uromyces appendiculatus,
Phytophthora palmivora and Phytophthora infestans (Shepherd et al., 2003; Ebstrup et al., 2005; Cooper et al.,
2006; Padliya and Cooper, 2006).
In the plant-pathogen interaction system, it is still very challenging to study proteome of pathogen in planta
since the biomass of the pathogen is a small portion of the total in the infected plant resulting in the dominance
Chapter 1 Introduction
15
of plant proteins. However, there is some achievement in the study of morphogenesis-, host- and signaling-
responsive protein and proteome mapping in fungal phytopathogens by proteomic techniques. M. grisea causing
rice blast disease, one of the most damaging diseases of rice, forms an appressorium from a germinating
conidium allowing the infection peg to penetrate the rice cuticle (Kim et al., 2004b). Due to the vital function of
the appressorium in disease initiation, gel-based proteomics was used to identify proteins during formation of the
appressorium, revealing five proteins including 2 α-subunits of the 20S proteasome, serine carboxypeptidase Y
and scytalone dehydratase (Kim et al., 2004b). The changes of the extracellular and intracellular proteomes of M.
grisea were examined when exposed to extracts from resistant and susceptible rice cultivars (Kachroo et al.,
1997). Protein spots induced by susceptible cultivar extracts were observed but not identified due to the limited
availability of gene sequences at the time of study. U. maydis is the causal agent of smut in corn which
undergoes a dimorphic transition from budding to form infective filamentous hypea (Bohmer et al., 2007). By
using 2-DE, 250 different proteins were identified from cells of the fungus U. maydis cultured in vitro. In
addition, in this study it was observed that 13 proteins involved in energy and general metabolism were
upregulated during the filamentous growth stage and signaling pathway involving a small GTP binding protein is
responsible for the generation of the filament during pathogenic development (Bohmer et al., 2007). The role of
signal transduction in the pathogenicity of S. nodorum which causes glume blotch in wheat is well established
(Tan et al., 2008). The Gna1 protein of cAMP pathway is of particular interest, as mutants displayed multiple
phenotypic impairments such as reduced virulence, reduced extracellular depolymerase activities and abolished
asexual sporulation (Tan et al., 2009). In order to identify the Gna1-regulated proteins, comparison of 2D
proteome patterns of wildtype and gna1 mutant coupled with LC-MS/MS was conducted, resulting in
identification of positively regulated short-chain dehydrogenase which has critical roles in asexual sporulation
and mycotoxin production (Tan et al., 2008, 2009). Cooper et al. (2006, 2007) has identified 468 and 461
proteins from uredospores and germlings, respectively, of U. appendiculatus, the rust fungus occurring on beans
by LC-MS/MS. Both proteomes contained proteins involved in protein biosynthesis and folding, suggesting that
spores and germlings become metabolically active primed by protein accumulation during infection. The
proteome maps of B. cinerea mainly infecting wine grapes and S. sclerotiorum causing a disease called white
mold in many flowers and vegetables were established (Fernandez-Acero et al., 2006; Yajima and Kav, 2006). A
comparison of the mycelial protein profiles of B. cinerea strains differing in toxin production revealed
differentially expressed malate dehydrogenase and glyceraldehyde-3-phosphate dehydrogenase between stains
(Fernandez-Acero et al., 2006). The mycelial proteome and the secretome of S. sclerotiorum were analysed,
resulting in identification of approximately 100 mycelial protein and 18 secreted proteins including cell-wall-
degrading enzymes (Yajima and Kav, 2006). There are some proteomic studies on F. graminearum. It was
shown by 2-DE and MS that the in vitro exoproteome of F. graminearum grown on glucose and on hop cell
Chapter 1 Introduction
16
walls contained 23 and 84 unique proteins, respectively, mainly involved in cell wall polysaccharide degradation
(Phalip et al., 2005). By high-throughput LC-MS/MS, 229 and 120 fungal proteins, mainly including glycoside
hydrolases and proteases, were identified in the secretome of F. graminearum during growth on 13 synthetic
media with carbon supplements and during infection of wheat heads, respectively (Paper et al., 2007). A gel-
based proteomic approach was employed to identify F. graminearum proteins secreted to culture medium
containing barley or wheat grain flour, revealing 155 fungal protein identifications in 69 unique proteins in either
medium which mainly included enzymes involved in degradation of cell walls, starch and proteins (Yang,
unpublished data, see Chapter 4).
With regard to the plant response to pathogens, it has been found that proteins involved in diverse biological
processes including defense and stress response, signal transduction, photosynthesis, protein folding and
degradation and energy metabolism are regulated (Thurston et al., 2005; Mehta et al., 2008). Some examples
reporting these proteins are mentioned here. The M. grisea-rice interaction has been well studied because of its
great economic importance and availability of both genome sequences (Mehta et al., 2008). Gel-based proteomic
analysis of rice leaves including the resistant and susceptible lines infected by M. grisea showed the induction of
two receptor-like protein kinases, two β-1,3-glucanases, thaumatin-like protein, peroxidase, probenazole-
inducible protein and rice PR10 protein in both lines (Kim et al., 2004a). Callose deposition and hypersensitive
response was clearly visible in incompatible interactions but excessive invading hypha with branches was
evident only in compatible interactions in this study (Kim et al., 2004a). It was reported that susceptibility of rice
to rice blast disease increased with the excessive application of nitrogen nutrients (Long et al., 2000). Therefore,
a study about effect of nitrogen nutrients on rice blast disease by proteomic approach was conducted, suggesting
that proteins involved in photosynthesis was affected in the interaction and twelve proteins changed in response
to different levels of nitrogen nutrient. Among these proteins, level of ribulose-1,5-bisphosphate
carboxylase/oxygenase was increased with higher level of N (Konishi et al., 2001). Protein profiles of blackleg-
resistant and susceptible canola cultivars after inoculation with Leptosphaeria maculans were investigated using
2-DE and tandem MS. Several antioxidant enzymes, including dehydroascorbate reductase and peroxiredoxin
along with proteins involved in photosynthetic and nitrogen metabolism were found to be upregulated in the
resistant cultivar compared to the susceptible cultivar (Subramanian et al., 2005). Different Fusarium species
can cause different diseases in a diversity of plant hosts. Proteome analysis of the xylem sap of tomato in
response to Fusarium oxysporum infection revealed accumulation of PR proteins such as glucanases,
peroxidases and chitinases, polygalacturonase and a subtilisin-like protease, which were involved in defense,
antioxidant protection and cell structure, as well as seven fungal proteins including arabinanase, oxidoreductase
and serine protease (Rep et al., 2002; Houterman et al., 2007). Gel-based proteomics was performed to study
the changes in the protein profiles of germinating maize embryos following infection by Fusarium verticillioides,
Chapter 1 Introduction
17
leading to the identification of PR proteins, antioxidant enzymes and protein involved in protein synthesis,
folding and stabilization (Campo et al., 2004). Several proteome analysis of barley and wheat in response to
Fusarium graminearum infection showed the induction of plant proteins associated with oxidative stress or
pathogenesis-related responses and changes of abundance of the proteins involved in primary metabolism and
protein synthesis (Zhou et al., 2006; Geddes et al., 2008; Yang et al., 2010b). In addition, transcriptome and
metabolome analysis have been performed in Fusarium graminearum–barley or wheat interaction to gain more
insights into the plant defense response to this pathogen. Microarray analysis of Fusarium graminearum-
Trail F. For blighted waves of grain: Fusarium graminearum in the postgenomics era. Plant Physiol. 2009, 149,
103–110.
Unlu M, Morgan ME, Minden JS. Difference gel electrophoresis: a single gel method for detecting changes in
protein extracts. Electrophoresis 1997, 18, 2071–2077.
van Loon LC, Rep M, Pieterse CMJ. Significance of inducible defense related proteins in infected plants. Ann.
Rev. Phytopathol. 2006, 44, 1–28.
van Loon LC, van Strien EA. The families of pathogenesis-related proteins, their activities, and comparative
analysis of PR-1 type proteins. Physiol. Mol. Plant Pathol. 1999, 55, 85–97.
Walter S, Nicholson P, Doohan FM. Action and reaction of host and pathogen during Fusarium head blight
disease. New Phytol. 2009, 185, 54–66.
Yajima W, Kav NN. The proteome of the phytopathogenic fungus Sclerotinia sclerotiorum. Proteomics 2006, 6,
5995–6007.
Yang F, Jensen JD, Spliid NH, Svensson B, Jacobsen S, Jørgensen LN, Jørgensen HJ, Collinge DB, Finnie C.
Investigation of the effect of nitrogen on severity of Fusarium head blight in barley. J. Proteomics 2010a, 73,
743–752.
Yang F, Jensen JD, Svensson B, Jørgensen HJ, Collinge DB, Finnie C. Analysis of early events in the interaction
between Fusarium graminearum and the susceptible barley (Hordeum vulgare) cultivar Scarlett. Proteomics
2010b, 10, 3748–3755.
Yang F, Jensen JD, Svensson B, Jørgensen HJ, Collinge DB, Finnie C. Secretome-based proteomics for
uncovering pathogenicity factors in Fusarium graminearum during interaction with barley and wheat. For
submission to Mol. Plant Pathol.
Zhou WC, Eudes F, Laroche A. Identification of differentially regulated proteins in response to a compatible
interaction between the pathogen Fusarium graminearum and its host, Triticum aestivum. Proteomics 2006, 6,
4599–4609.
Chapter 2
27
Chapter 2
Investigation of the effect of nitrogen on severity of
Fusarium Head Blight in barley
J O U R N A L O F P R O T E O M I C S 7 3 ( 2 0 1 0 ) 7 4 3 – 7 5 2
ava i l ab l e a t www.sc i enced i r ec t . com
www.e l sev i e r . com/ loca te / j p ro t
Investigation of the effect of nitrogen on severity of FusariumHead Blight in barley
Fen Yanga, Jens D. Jensenb, Niels Henrik Spliidc, Birte Svenssona, Susanne Jacobsena,Lise Nistrup Jørgensenc, Hans J.L. Jørgensenb, David B. Collingeb, Christine Finniea,⁎a Enzyme and Protein Chemistry, Department of Systems Biology, Technical University of Denmark, Denmarkb Department of Plant Biology and Biotechnology, Faculty of Life Sciences, University of Copenhagen, Denmarkc Department of Integrated Pest Management, Faculty of Agricultural Sciences, Aarhus University, Denmark
A R T I C L E I N F O
Abbreviations: BBBI, Barley Bowman-Birk inFHB, Fusarium Head Blight; NIV, Nivalenol; PC⁎ Corresponding author. Enzyme and Protein C
Article history:Received 10 September 2009Accepted 23 October 2009
The effect of nitrogen on Fusarium Head Blight (FHB) in a susceptible barley cultivar wasinvestigated using gel-based proteomics. Barley grown with either 15 or 100 kg ha−1Nfertilizer was inoculated with Fusarium graminearum (Fg). The storage protein fraction did notchange significantly in response either to N level or Fg, whereas eighty protein spots in thewater-soluble albumin fraction increased and 108 spots decreased more than two-fold inintensity in response to Fg. Spots with greater intensity in infected plants contained fungalproteins (9 spots) and proteolytic fragments of plant proteins (65 spots). Identified fungalproteins included two superoxide dismutases, L-xylulose reductase in two spots, peptidylprolyl cis–trans isomerase and triosephosphate isomerase, and proteins of unknownfunction. Spots decreasing in intensity in response to Fg contained plant proteins possiblydegraded by fungal proteases. Greater spot volume changes occurred in response to Fg inplants grown with low nitrogen, although proteomes of uninfected plants were similar forboth treatments. Correlation of proteome changes withmeasurement of Fusarium-damagedkernels, fungal biomass and mycotoxin levels indicated that increased Fusarium infectionoccurred in barley with low N and suggests control of N fertilization as a possible way tominimise FHB in barley.
Fusarium Head Blight (FHB) or scab, caused by Fusariumspecies including Fusarium graminearum (Fg) Schwabe (tele-omorph: Gibberella zeae (Schwein) Petch) in humid andsemihumid climates, is a devastating disease in wheat(Triticum aestivum), barley (Hordeum vulgare) and other cerealsand has the capacity to destroy a potentially high yield [1–3].The disease reduces the grain yield due to floret sterility as
hibitor; DON, DeoxynivalA, Principal component ahemistry, Department of. Tel.: +45 4525 2739; fax:
er B.V. All rights reserved
well as poor grain filling and reduced kernel size [4]. In additionto deceased yield and quality, the infected grains often containmycotoxins like deoxynivalenol (DON), nivalenol (NIV) andzearalenone (ZEA) which are hazardous to animals andhumans causing neurological disorders and immunosuppres-sion due to inhibition of protein biosynthesis [5].
The pathogen Fg attacks the barley spikes after theyemerge from the flag leaf sheath in the late-milk to soft-dough stages of seed development [6]. Disease symptoms
744 J O U R N A L O F P R O T E O M I C S 7 3 ( 2 0 1 0 ) 7 4 3 – 7 5 2
include premature necrosis and a brown/grey discolourationof spike tissue. In contrast to wheat, where the fungus spreadsbetween spikes of the ear through the rachis, the fungus doesnot spread in barley [2,7].
It is highly challenging to control FHB in barley and wheatdue to the poor understanding of the mechanisms of plantresistance, and since no highly resistant barley cultivar is yetavailable [2]. Development of FHB-resistant cultivars is a high-priority for many barley breeding programs worldwide.Agronomic and crop management strategies aimed at con-trolling FHB in wheat and barley include foliar fungicideapplication, crop rotation and tillage practices, however theseare in general not highly effective [8–10]. Environmentalfactors play an important role in pathogenesis. High temper-ature and humidity levels (e.g. heavy dew) favour fungalattack and disease development [11,12].
Several studies, all carried out in Canada, suggest that thetype and amount of nitrogen fertilizer can affect the incidenceand severity of FHB, however the observed effect on thedisease differed. Thus, in wheat and barley grown with aninitial application of 70 kg ha−1N (ammonium nitrate) atseedling stage, a significant increase of FHB was observed onplants supplied with an additional 50 kg ha−1 at Zadoks'growth stages 30 and 45, in comparison with plants receivingno supplement [10]. On the other hand, wheat grown in fieldswith 90 kg ha−1N (ammonium nitrate, urea or both) or lessshowed significantly higher levels of FHB than those withmore N [13]. It was also reported that wheat grown in clayloam or sandy loam soil without application of N had a greaterincidence of FHB than with 100 kg ha−1N (ammonium nitrate)[4]. In maize, soil N amendment with 100 kg ha−1 ammoniumnitrate decreased both disease severity and levels of DON [14].It is clear from these results that more research is required todetermine how the N status of the host plant may influencethe interaction with Fg.
Transcriptome profiling using microarrays, metabolomeprofiling by aid of GC/MS and proteome profiling have beenused to investigate plant responses during the infection by Fg[2,3,7,15,16]. Microarray analysis of the barley–Fg interactionshowed induction of plant genes encoding defence responseproteins, oxidative burst-associated enzymes, phenylpropa-noid pathway enzymes, and trichothecene and tryptophancatabolic enzymes [7]. Metabolome analysis of wheat infectedby Fg indicated the higher abundance of several fatty acidsand aromatic compounds in both susceptible and relativelyresistant cultivars after infection whereas coumaric acids,myo-inositol, certain sugars andmalonic acidwere only foundin the relatively resistant cultivar [15]. Gel-based proteomeanalysis of barley and wheat in response to Fg infectionindicated that plant proteins associated with oxidative stressor pathogenesis-related responses were induced whereasproteins involved in photosynthesis and carbon metabolismdecreased in abundance [2,3,16]. Additionally, proteomeanalysis was applied to Fg grown in vitro on synthetic mediaand in planta during infection of wheat head. In that study, 120fungal proteins including secreted proteins and housekeepingenzymes were identified by LC-MS/MS in planta [17].
Proteomics is a useful approach for studying plant–pathogen interactions as differentially expressed proteinsdirectly involved in plant–pathogen responses can be detected
by comparing protein profiles [2,16,18]. Barley seed proteomesand their genetic and developmental variations have beendescribed in some detail [19] and include the identification ofseveral hundred proteins by mass spectrometry. This knowl-edge provides a solid background on which to base aninvestigation of the combined effect of N fertilization andFHB on the barley seed proteome.
In the present study, 2-DE followed by MALDI-TOF massspectrometry was employed to examine changes in theproteome of mature barley seeds caused by Fg infectionunder different levels of N fertilization, providing the firstmolecular insight into the effect of N on FHB infection ofbarley. The proteome changes were correlated with quantifi-cation of mycotoxins and fungal biomass and includedidentification of some fungal proteins not previously observedThe results suggest that, even though the difference in
nitrogen level does not appear to greatly affect abundance orcomposition of grain albumins and storage proteins, theseverity of Fg infection increases significantly in plantsgrown with low N.
2. Materials and methods
2.1. Plant growth
Barley (cv. Scarlett) from a single batch was grown in plasticpots containing soil:sand:peat 2:1:1 (w/w/w) and necessarymicro- and macronutrients with or without addition ofnitrogen fertilizer (ammonium nitrate) resulting in nitrogenlevels of around 100 (high N) or 15 kg ha−1 (low N). Aftergermination, excess seedlings were removed and twelveseedlings per 8-litre pot (25 cm diameter) were grown tomaturity in an unheated greenhouse under natural lightconditions at 15–22 °C during the day (12 h) and 10–15 °Cduring the night (12 h). Sufficient water was supplied using adripwatering system. The leafwater potentialwas determinedusing a pressure chamber andwateringwas adjusted to a levelbetween0 and−0.5 mPa. Three biological replicates, consistingof one pot each, were prepared for Fg and control inoculationsunder low and high nitrogen levels.
2.2. Inoculum preparation and spike inoculation
F. graminearum R-77550 was grown on potato-dextrose agar(PDA) for 2 weeks before inoculation ofmung bean broth (MBB)for macroconidia production. MBB was made by mixing 1 Lboiling water with 40 g of mung bean seeds for 10 minfollowed by filtration through two layers of cheesecloth toremove seeds before autoclaving. MBB was then inoculatedwith 10 plugs (0.5 cm) of the Fg PDA culture and incubated on ashaker (200 rpm) at 26 °C in darkness for 4 days. Macroconidiawere harvested by filtering the culture through two layers ofcheesecloth to remove mycelium. The concentration ofmacroconidia was adjusted to 5×104spores/mL with deio-nised water containing 0.1% Tween 20. Inoculation of barleyspikes was conducted by applying 5 mL of inoculum at theanthesis stage 65 [20] using a “handsprayer”. Control plantswere mock-inoculated with water. After inoculation, spikeswere kept under sealed plastic for 72 h. At least five spikes
745J O U R N A L O F P R O T E O M I C S 7 3 ( 2 0 1 0 ) 7 4 3 – 7 5 2
were harvested at maturity stage for each replicate. Thepercentage of Fusarium-damaged kernels (FDK) was deter-mined based on kernel colour and degree of shrivelling foreach grain [21].
2.3. F. graminearum biomass determination
Kernels were ground in a cooled mill (4 °C) for 30 s and theflour was used for fungal biomass determination, mycotoxinanalysis and protein extraction. Fg DNA was extracted from100 mg of flour using a CTAB protocol followed by partialpurification of DNA on a Qiagen DNeasy plant mini kit asdescribed [22]. Samples were diluted to a concentration of10 ng/nL total DNA before PCR. Primers for Fg elongation factor1α (FusEF14 forward: 5′-ccacgtcgactctggcaag and FusEF125reverse: 5′-cgcactggtagatcaagtgacc) were obtained from M.Nicolaisen, University of Aarhus, DJF, Flakkebjerg, Denmark.Both primer pairs were initially tested in the Mx3000P real-time PCR machine (Stratagene) by making standard anddissociation curves on a Fg DNA dilution series ranging from10 ng to 1 pg, giving a detection limit of about 25 nuclei. PCRwas performed with 1 µL of template DNA, 10 pmol of eachforward and reverse primer, 12.5 µL of 2× SYBR Green mastermix (Applied Biosystems) and 0.4 µL of a 500× dilutedreference dye (Applied Biosystems) in a final volume of 25 µLand using the following program 95 °C for 10 min, 40 cycles of15 s at 95 °C, 30 s at 60 °C and 30 s at 72 °C. Fluorescence wasdetected after each cycle. After the last amplification cycle, thespecificity of the PCR was determined in a melting curveanalysis by increasing the temperature from 60 °C to 95 °Cwhile measuring the fluorescence for every 0.5 °C increase. Ano-template control was run with the primer pair.
2.4. Mycotoxin analysis
The mycotoxin analysis was performed as described [23] withslight modifications. One gram of milled grain was extractedwith 4 mL of acetonitrile:water [84:16 (v/v)], followed bycentrifugation. Two hundred microlitres of the supernatantwas mixed with 700 µL water and 100 µL internal standard(13C-labelled DON in 25% acetonitrile/water). The sampleswere filtered through a 0.45 µm filter before analysis by LC-MS/MS. The chromatographic separation was performed with aHewlett-Packard 1100 system with gradient elution. Fortymicrolitres was injected onto a 250×2.1 mm BDS Hypersil C18.5 µm column (Thermo Electron Corporation, Waltham,Massachusetts, US). The A-eluent was 99% water and 1%methanol and the B-eluent was 10% water and 90% methanol.MS/MS detection was carried out using a Sciex API 2000instrument (Applied Biosystems) in electrospray negativemultiple reaction ionisation mode for DON, NIV and ZEAtoxins. Detection limits were 10 µg kg−1 for DON and NIV, and2 µg kg−1 for ZEA. Relative standard deviations were 10%.
2.5. Protein extraction and quantification
Water-soluble protein was extracted as described [24] withminor modifications. Flour (200 mg) was extracted with 1 mL5 mM Tris–HCl, pH 7.5, 1 mM CaCl2 containing the proteaseinhibitor cocktail “complete” (Roche) by shaking (Eppendorf
Thermomixer, maximum speed) for 30 min at 4 °C [25].Duplicate extractions were carried out for each biologicalreplicate such that six protein extracts were prepared pertreatment. After centrifugation (20,000×g×30 min, 4 °C), thesupernatant was stored in aliquots at −80 °C until needed. Theprotein concentration was determined by the Popov AmidoBlack-based method [26] with bovine serum albumin asstandard. About 8 mg protein was obtained per gram flour.After extraction of the soluble protein fraction, the flourpellets were re-extracted with 800 µL of 70% ethanol asdescribed [27] to obtain the storage protein fraction. As thePopov method is not ideal for determination of the concen-tration of alcohol-soluble proteins, the protein concentrationof storage protein fractions was determined using theBradford assay [28]. Storage protein extractions yielded about1.5 mg protein/g flour.
2.6. Two-dimensional gel electrophoresis
For 2-DE of water-soluble protein extracts, 200 µg protein wasprecipitated by 4 volumes of acetone at −20 °C overnight.Protein was dissolved in 350 µL of “reswelling” buffer contain-ing 7 M urea, 2 M thiourea, 2% (w/v) CHAPS, 0.5% (v/v) IPGampholytes (pH 4–7), “Destreak reagent” 1.2% (v/v) (GEHealthcare) and a trace of bromophenol blue. The proteaseinhibitor cocktail “complete” (Roche) was added to avoidproteolysis during the isoelectric focusing step [25]. Sampleswere centrifuged before application to 18 cm pI 4–7 IPG strips.Isoelectric focusing (IEF) was run on an Ettan IPGphor (GEHealthcare) for a total of 60,000 Vh as described [24]. Seconddimension SDS-PAGE (12–14%, 18 cm×24 cm, GE Healthcare)was performed on a Multiphor II (GE Healthcare) as described[24]. For 2-DE of storage protein extracts, 50 µg protein wasvacuum-dried before dissolving in 350 µL of “reswelling”buffer as above. Gels were stained by colloidal CoomassieBrilliant Blue [29].
2.7. Image analysis
Scanned gel images (greyscale, 16bit) were imported into theimage analysis software Progenesis SameSpots (NonlinearDynamics, UK). All gel images were warped, matched andaligned to a chosen reference gel. Six or four imagesrepresenting three or two biological and two technicalreplicates for each of the four treatments were grouped togain the average volume of each spot. The protein spotvolumes were automatically normalised in the software. Alist of spots which changed in abundance among the fourtreatments was generated. A threshold of ANOVA (p)<0.05,q<0.05, power >0.8 and at least two-fold change in averagespot volume was used to define the protein spots chosen forfurther analysis. Principal component analysis (PCA) in theimage analysis software was applied to analyse the similarityof protein patterns among gels and the expression profiles ofprotein spots fulfilling the above criteria.
2.8. In-gel digestion
Spots were excised from gels and subjected to in-gel trypsindigestion as described [30] with minor modifications. Gel pieces
746 J O U R N A L O F P R O T E O M I C S 7 3 ( 2 0 1 0 ) 7 4 3 – 7 5 2
werewashedwith 40% ethanol, shrunk by 100% acetonitrile andsoaked in 2 µL 11.1 ng/µL trypsin (Promega, porcine sequencinggrade) in 25mM NH4HCO3 on ice for 45 min. Six µL of 25 mMNH4HCO3wasadded to gel pieces followedby incubationat 37 °Covernight. Sampleswereprepared forMALDI-TOFanalysisonanAnchorchip™ Target (Bruker-Daltonics, Bremen, Germany) as
described [30]. A tryptic digest of β-lactoglobulin was used forexternal calibration.
2.9. Protein identification
An Ultraflex II mass spectrometer (Bruker-Daltonics, Bremen,Germany) was used for peptide mass mapping or peptidefragment ion mapping. An in-house Mascot server (http://www.matrixscience.com) was used for database searches inthe NCBInr at the National Center for Biotechnology Informa-tion, the HvGI barley gene index Release 10.0 (http://compbio.dfci.harvard.edu/tgi) and the Broad Institute for F. graminearumgene index (http://www.broad.mit.edu/annotation/genome/fusarium_graminearum). The following parameters were usedfor searching: allowed global modification, carbamidomethylcysteine; variable modification, oxidation of methionine;missed cleavages, 1; peptide tolerance, 80 ppm and MS/MStolerance ±0.5 Da. To be considered asapositive identification, asignificant score calculated by the Mowse scoring algorithm inMASCOT(above67, 69and54 respectively for theHvGI,NCBIandBroad Institute databases) was required with at least fourmatched independent peptides for peptide mass mappingor two matched peptides in MS/MS analysis. All identifiedfungal protein sequences were assessed for the presence ofsignal peptides using SignalP (http://www.cbs.dtu.dk/services/SignalP), andsequencesencodingproteinsofunknown functionwere subjected to BLAST search in NCBI.
3. Results
3.1. Disease incidence
Barley plants grown with low or high N were inoculated withFg or water as control. The percentage of Fusarium-damagedkernels (FDK) was higher for inoculated plants grownwith lowN than with high N (Fig. 1A). Since it can be difficult todistinguish FDK from other types of kernel damage ordiscolouration, fungal biomass and mycotoxin levels in thesampleswere alsomeasured to determine the degree of fungalinfection more precisely.
In agreement with the FDK analysis, the concentration ofFg DNA was very low or undetectable in control plants andhigher in Fg-infected plants grown with low N than with highN (Fig. 1B). DON is the main mycotoxin produced by Fg, andagain there was a clear increase in DON in infected samplestreated with low N with respect to high N (Fig. 1C). Only verylow levels of NIV (Fig. 1D) and ZEA (Fig. 1E) could be detected.Low levels of FgDNA andmycotoxins in some control samplesindicated low contamination from the natural environment.
Fig. 1 – Incidence of Fusarium infection in barley cv. Scarlettgrown under low and high N. A. Percentage ofFusarium-damaged kernels (FDK); B. Fungal biomass expressedas concentration of Fg DNA; C. Concentration of FgmycotoxinDON; D. Concentration of Fgmycotoxin NIV; E. Concentration ofFgmycotoxin ZEA. The three biological replicates are shown foreach treatment. In some cases, indicated by #, measurementswere below the detection limit.
747J O U R N A L O F P R O T E O M I C S 7 3 ( 2 0 1 0 ) 7 4 3 – 7 5 2
When comparing the three biological replicates, it wasobserved that both Fg DNA content and toxin concentrationwere much lower in biological replicate 1 from Fg-infected,low N than in replicates 2 and 3 (Fig. 1).
3.2. Proteome analysis
Two-dimensionalgelelectrophoresis (pH4–7)wasrunusing threebiological replicates and two technical replicates fromeach of thefour treatments. The 2-DE patterns of the non-infected samplesgrown with high and low N were highly similar to each other.Clear differences were apparent between Fg-infected and controlsamples grown under lowNbutwere less apparent under highN(Fig. 2A). In particular, the intensity of several high molecularweight proteins decreased and many new spots with lowermolecular weight appeared in the Fg-infected extracts (Fig. 2). Inagreement with the FDK, fungal DNA and toxin analysis (Fig. 1),the 2-DE pattern of biological replicate 1 from Fg-infected, low Nwas intermediate between replicates 2 and 3 and the Fg-infected,high N samples when the gel images were subjected to principal
Fig. 2 – Water-soluble protein profiles of barley seeds. A. One reprerange 4–7 is shown for each treatment.Molecular sizemarkers are ishowing differences between infected and control treatments; B. C
component analysis (PCA; data not shown). This replicate wastherefore removed from the subsequent analysis of 2-DE spotvolumes. The statistical analysis was thus based on 22 gels. Intotal, 466 protein spots were detected on all gel images based onthe criteria (ANOVA<0.05, q<0.05 and power >0.8). PCAwas usedto examine the relationship of the individual spots with the 2-DEpatterns (Fig. 3A). The biological and technical replicates fromeach treatment clustered together, demonstrating the reproduc-ibility of the 2-DE patterns. Sixty percent of the variance could beexplained by principal component 1, which separated the gelsaccording to the degree of Fg infection of the samples, whereasonly about 6% of the variance was explained by principalcomponent 2, separating gels based on the N level. The greatestdifference was observed between the Fg-infected and controlsamples grown with low N in agreement with the FDK, Fgbiomass and toxin data (Fig. 1).
The average volumes of 188 spots varied by at least two-foldamong the four treatments. These spotswere chosen for furtheranalysis. The spots formed two clusters according to theirappearance profiles (Fig. 3B). Protein spots in cluster A (80 spots)
sentative colloidal Coomassie blue-stained gel covering the pHndicated. Numbered boxes indicate selected regions of 2D-gelslose-up views of boxes for infected and control low N samples.
Fig. 3 – Protein spots varying in response to Fg infection and N level. A. Principal component analysis. Protein spots showingsignificant variation are indicated by numbers and gel images are indicated by circles. Biological and technical replicates aregrouped in circles corresponding to the four treatments. B. Expression profiles of protein spots in cluster A (increased intensitywith Fg infection) and cluster B (decreased intensity with Fg infection). Each trace corresponds to one 2-DE spot and each nodecorresponds to one replicate. Functional categories of the plant proteins identified in clusters A and B are indicated.
748 J O U R N A L O F P R O T E O M I C S 7 3 ( 2 0 1 0 ) 7 4 3 – 7 5 2
increased and spots in cluster B (108 spots) decreased inintensity in response to Fg. For each spot, a greater change inintensity in response to Fgwas observed for lowN than for highN except spots 119, 238 and 263 from cluster A which increasedmore in intensity under high N. No spots showed greater than1.9-fold changewhen comparing theuninfected samples grownwith low and high N (data not shown), supporting previousfindings that altered levels ofN fertilizerdonot greatly affect thewater-soluble seed proteome, although some individual pro-teins may be differentially expressed [31]. Overall, the resultssuggest that the observed changes in protein patterns arecaused mainly by the Fg infection, and that these changes aregreater in plants grown with lower levels of nitrogen.
In cluster A, containing the spots increasing in intensity inresponse to Fg, 9 spots were identified as fungal proteins,
whereas 67 barley proteins were identified in 65 spots. Sixspots could not be identified. Based on sequence coverageobtained in MS and lower observed molecular weight thanexpected (Supplementary Table S1), as well as previousidentification of the same proteins in spots of expectedmolecular weight [32], most of the identified plant proteinswere judged to be proteolytic fragments (Supplementary TableS1, Fig. S1). In cluster B, 109 barley proteins were identified in107 spots (Supplementary Table S2, Fig. S2). A single spot couldnot be identified.
The identified plant proteins were assigned to functionalcategories (Fig. 3B). Plant proteins of known function whichdecreased in response to Fg infection were mainly chaperones,defence-related proteins, proteins involved in desiccation andoxidative stress, proteins involved in primary metabolism such
749J O U R N A L O F P R O T E O M I C S 7 3 ( 2 0 1 0 ) 7 4 3 – 7 5 2
as glycolysis, starchmetabolism, citric acid cycle and amino acidbiosynthesis, the volume of which changed between 7.7-fold(Spot 1; protein similar to protamine P1; Supplementary Fig. S2)and 2-fold (Spots 350,379,395; serpin; Supplementary Fig. S2). Theplant proteins in the spots increasing in intensity in response toFg infection belonged to similar functional categories as theproteins that decreased in abundance, suggesting that plantproteins in cluster B were degraded to produce the fragments incluster A. The main degradation products in cluster A weregenerated frommajor seedwater-solubleproteins suchas serpin,β-amylase and low molecular weight α-amylase and proteaseinhibitors likeCI-1A,CI-1B,CI-2A,BMAIandBDAI.However, otherspots previously identified as α-amylase/trypsin inhibitors CMa,CMb, CMd and trypsin/amylase inhibitor pUP 13 [32] were foundwith slightly increased or unchanged intensity (data not shown).
Nine fungal proteins were identified, suggesting that thesewere relatively abundant in the fungal in planta proteome(Table 1). Six of these spots were identified as proteins ofannotated function. The remaining three fungal protein spotswere annotated as hypothetical proteins. None of the fungalproteins were predicted to have signal peptides.
Analysis of the storage protein fractionswas also carried out,but no spot volume change greater than 2-fold was observed inthese fractions (data not shown).
4. Discussion
4.1. The effect of nitrogen fertilization on the proteome ofinfected grain
Analysis of FDK, fungal biomass and mycotoxin levels clearlyshowed that Fg infection was more severe in barley plantsgrown under low N compared to high N, suggesting that highFHB incidence correlates with low N status. Nitrogen canincrease biomass of leaf and stalk tissue and enable highergrain yield [33] as well as affect plant responses to pathogens[14,33]. Furthermore, it has been reported that nitrogendepletion in the medium used for in vitro fungal growth
Table 1 – Fusarium graminearum proteins identified in infected
a Spot numbers refer to Supplementary Fig. S1.b FGSG: accession number in Fusarium graminearum gene index from Broac Protein also identified in [17].d Two peptides additionally confirmed by MS/MS analysis.e Protein also identified in [3].
could induce synthesis of trichothecene [34,35] and depletionof nitrogen or nutrients was required for Fusarium spores togerminate and infect plants [33,36]. However, in the presentstudy, although high N plants displayed delayed floweringtime and prolonged grain filling, no major difference wasobserved in the appearance of the control plants under twonitrogen levels at maturity stage. Since the ratio of DON to FgDNA was similar under low and high N, it seems unlikely thatthe mycotoxin is produced as a response to low N status.
With respect to the analysis of the water-soluble proteome,we observed similar protein profiles of the two uninoculatedgroups and clear differences between infected and controlgels, where considerable differences in response to fungalinfection occurredwith lowN, indicating that incidence of FHBdisease in barley grain was increased. The proteome resultswere in good agreement with analysis of FDK, fungal biomassand mycotoxin levels.
The group of proteins down-regulated in response to Fginfection include functions such as chaperones, defence-related proteins, proteins related to desiccation and oxidativestress and proteins involved in primary metabolism. Partial orcomplete disappearance of the protein matrix and starchgranules has been reported in kernels of spring wheatdamaged by Fusarium using scanning electron microscopy[37]. Protein bodies of wheat kernels were degraded topolypeptides of lower molecular weight during infection byFg [38,39]. An increase in protease activity of diseased barleygrain has also been shown [40]. Fusarium species werediscovered to produce alkaline proteases associated with thedegradation of certain water-soluble proteins in infestedbarley [41]. Several Fg peptidases and aspartyl proteaseswere also identified during infection of wheat head [17]. It ispossible that in our study, the observed widespread proteolysiswas caused by fungal proteases acting as pathogenicity factors,produced to obtain nutrition from the host [42,43]. Whilefragments of barley protease inhibitors such as CI-1A, CI-1B,CI-2A, BMAIandBDAIwere observed, others suchasα-amylase/trypsin inhibitors were not reduced in amount, suggestingdifferential sensitivity to proteases. Previous studies performed
750 J O U R N A L O F P R O T E O M I C S 7 3 ( 2 0 1 0 ) 7 4 3 – 7 5 2
on the barley protease inhibitors CI-1A, CI-1B, CI-2A, BASI andbarleyBowman-Birk inhibitor (BBBI) haveshownthat theydifferin their ability to inhibit Fusarium serine proteases [44,45]. Theinterplay between fungal proteases and plant protease inhibi-tors may thus influence the outcome of the host–pathogeninteraction [45].
Alternatively, the observed degradation may be a conse-quence of endogenous proteases as a result of necrosis ordefence responses. Exactly which proteases contribute to thedegradation of albumins in vivo remains unclear since noproteases were identified in the infected seed proteomes.Further investigation of the proteases causing hydrolysis ofthe barley proteome in vivo is required.
4.2. F. graminearum does not affect the storage proteinfraction
In contrast to the changes observed in the water-soluble seedproteome, no significant alteration in 2-DE patterns of thehordein storage protein fractionwas observed in response toNand/or Fg infection although it has been reported previouslythat higher N can result in an increase in hordein, whereas theother protein fractions including the water-soluble albuminfraction are relatively unaffected [46]. Our results indicatedthat the low nitrogen conditions in this study caused very fewchanges in storage protein or albumin fractions and that Fg-induced proteolysis mainly occurred in the albumin fractionregardless of fertilizer level. Fg may preferentially proteolysethe more accessible albumin fraction to establish infection,rather than the proline-rich storage proteins in the starchyendosperm. In agreement with this hypothesis, it was shownthat a major reduction of Fg and DON in diseased wheatkernels was achieved by pearling, a process in which most ofthe outer layers, including the aleurone layer, are removed [47]suggesting that the fungus has limited contact with thestorage proteins of the starchy endosperm. Our results suggestthat the observed effect of differential N levels on Fg infectionis not caused by a change in overall protein composition oramount in the seeds.
4.3. Proteins of fungal origin
Nine proteins were identified that originated from Fg, thegenome of which is predicted to contain 13,332 genes (www.broad.mit.edu). The Fg proteins were probably dominated bythe host plant proteins given the high ratio of plant to fungalbiomass [18]. In agreementwith this, only 25 Fg transcripts weredetected as accumulating 72h after inoculation of barley [7] andonly 8 fungal proteins were identified when the wheat–Fginteraction was examined using 2-DE [3]. In contrast, 120 inplanta Fg proteins were identified in vacuum-infiltrated fluidfrom infectedwheat heads [17], including five out of the 9 fungalproteins identified in the present study. However, the fact thatfour additional proteins were identified here strongly suggeststhat complementary approaches are required to analyse the inplanta fungal proteome. The five proteins common to thesestudies were Cu–Zn superoxide dismutase, triose phosphateisomerase, peptidyl-prolyl cis–trans isomerase and L-xylulosereductase (Table 1). The Cu–Zn superoxide dismutase andpeptidyl-prolyl cis–trans isomerase were also observed in the
Fg-infected wheat grain proteome [3]. The identification of anadditional form of superoxide dismutase in the present studystrongly suggests that the pathogen is exposed to, and attemptsto overcome plant defence-related reactive oxygen species. L-Xylulose reductase is involved in the assimilation of L-arabinose, derived from breakdown of plant cell walls, intothe fungal pentose phosphate pathway [48] and was identifiedin two spots varying in pI, illustrating a strength of 2-DE basedstudies and suggesting that the protein may be post-transla-tionally modified. It was however not possible on the basis ofthe MS data to determine the nature of the modification.
Three of the fungal proteins identified for the first time inthis study had unknown functions. One protein shared 82%sequence identity with torulene oxygenase from Fusariumfujikuroi. Torulene oxygenase is involved in the carotenoidpigment biosynthetic pathway [49], the initial steps of whichare common to biosynthesis of gibberellin, which is producedby F. fujikuroi in large amounts in low nitrogen medium [50].Therefore, nitrogen depletion in our study possibly inducesthe higher expression of genes in this pathway and suggeststhat a role for gibberellins may be found in this interaction.
One of the identified Fg proteins (Spot 157) with unknownfunction appeared in the 2-DE pattern with a pI of 4.2 which issurprisingly lower than the predicted theoretical value of 9.2.(Table 1, Supplementary Fig. S1). The observed molecularmass (49 kDa) was also slightly higher than expected (40 kDa).Such large pI discrepancy suggests that the protein is post-translationally modified. The protein sequence contains apotential Asn-X-Thr N-glycosylation site and the regions ofthe protein not covered by peptide mass data are rich in Serand Thr, suggesting that the protein could be glycosylated,however in-depth analysis by mass spectrometry will berequired to confirm this.
5. Conclusion
Taken together, the positive correlation of FDK, fungalbiomass, mycotoxins and proteome changes observed in thisstudy strongly suggests that FHB is more severe in barleygrown with low N than with high N. To our knowledge, this isthe first report to clarify the effect of nitrogen on FHB in barleyusing proteomic approaches. The Fg-infected proteome pat-terns of barley seeds reveal degradation of the water-solublealbumin protein fraction and detection of fungal proteins withmetabolic and antioxidative functions. Further work willconfirm whether appropriate N amendments in the field candecrease both disease severity and mycotoxin accumulation.
Acknowledgements
We thank Birgit Andersen (DTU) for technical assistance andSidsel Kirkegaard and Helene Saltoft Kristjansen (AarhusUniversity) for sowing and sampling plant material. Thiswork is funded by the Directorate for Food, Fisheries and AgriBusiness (DFFE) grant #3304 FVFP 060678 “Fusarium diseaseresistance— toxins and feed quality”, Plant Biotech Denmark,the Centre for Advanced Food Studies (LMC) and a PhD stipendfrom the Technical University of Denmark (DTU).
751J O U R N A L O F P R O T E O M I C S 7 3 ( 2 0 1 0 ) 7 4 3 – 7 5 2
Appendix A. Supplementary data
Supplementary data associated with this article can be found,in the online version, at doi:10.1016/j.jprot.2009.10.010.
R E F E R E N C E S
[1] McMullen M, Jones R, Gallenberg D. Scab of wheat and barley:a re-emerging disease of devastating impact. Plant Dis1997;81:1340–8.
[2] Geddes J, Eudes F, Laroche A, Selinger LB. Differentialexpression of proteins in response to the interaction betweenthe pathogen Fusarium graminearum and its host, Hordeumvulgare. Proteomics 2008;8:545–54.
[3] Zhou WC, Eudes F, Laroche A. Identification of differentiallyregulated proteins in response to a compatible interactionbetween the pathogen Fusarium graminearum and its host,Triticum aestivum. Proteomics 2006;6:4599–609.
[4] Subedi KD, Ma BL, Xue AG. Planting date and nitrogen effectson Fusarium Head Blight and leaf spotting diseases in springwheat. Agron J 2007;99:113–21.
[5] Goswami RS, Kistler HC. Heading for disaster: Fusariumgraminearum on cereal crops. Mol Plant Pathol 2004;5:515–25.
[6] BushnellWR, Hazen BE, Pritsch C. Histology and physiology ofFusarium Head Blight. In: Leonard KJ, Bushnell WR, editors.Fusarium Head Blight of wheat and barley. Saint Paul:American Phytopathological Society Press; 2003. p. 44–83.
[7] Boddu J, Cho S, Kruger WM, Muehlbauer GJ. Tanscriptomeanalysis of the barley-Fusarium graminearum interaction. MolPlant Microbe Interact 2006;19:407–17.
[8] Parry DW, Jenkinson P, McLeod L. Fusarium ear blight(scab) in small grain cereals: a review. Plant Pathol1995;44:207–38.
[9] Martin RA. Fusarium Head Blight in the Atlantic region, anoverview of R&D. Progress towards Fusarium Head Blightresistant spring wheat in Canada. Proc. 2nd CanadianWorkshop on Fusarium Head Blight, Ottawa, Canada. Ottawa:Agriculture and Agri-Food Canada; 2001. p. 114–5.
[10] Martin RA, MacLeod JA, Cladwell C. Influence of productioninputs on incidence of infection by Fusarium species of cerealseeds. Plant Dis 1991;75:784–8.
[11] Doohan FM, Brennan FM, Cooke BM. Influence of climaticfactors on Fusarium species pathogenic to cereals. Eur J PlantPathol 2003;109:755–68.
[12] Xu X. Effect of environmental conditions on the developmentof Fusarium Head Blight. Eur J Plant Pathol 2003;109:683–9.
[13] Teich AH, Nelson K. Survey of Fusarium Head Blight andpossible effects of cultural practices in wheat fields inLambton County in 1983. Can Plant Dis Surv 1984;64:11–3.
[14] Reid LM, Zhu X, Ma BL. Crop rotation and nitrogen effects onmaize susceptibility to gibberella (Fusarium graminearum) earrot. Plant Soil 2001;237:1–14.
[15] Hamzehzarghani H, Kushalappa AC, Dion Y, Rioux S, ComeauA, Yaylayan V, et al. Metabolic profiling and factor analysis todiscriminate quantitative resistance in wheat cultivarsagainst Fusarium Head Blight. Physiol Mol Plant Pathol2005;66:119–33.
[16] Wang Y, Yang LM, Xu HB, Li QF, Ma ZQ, Chu CG. Differentialproteomic analysis of proteins in wheat spikes induced byFusarium graminearum. Proteomics 2005;5:4496–503.
[17] Paper JM, Scott-Craig JS, Adhikari ND, Cuomo CA, Walton JD.Comparative proteomics of extracellular proteins in vitro andin planta from the pathogenic fungus Fusarium graminearum.Proteomics 2007;7:3171–83.
[18] Coiras M, Camafeita E, Lopez-Huertas MR, Calvo E, Lopez JA,Alcami J. Application of proteomics technology for analyzing
the interaction between host cells and intercellular infectiousagents. Proteomics 2008;8:852–73.
[19] Finnie C, Svensson B. Barley seed proteomics from spots tostructure. J Proteomics 2009;72:315–24.
[20] Zadoks JC, Chang TT, Konzak CF. A decimal code for thegrowth stages of cereals. Weed Res 1974;14:415–21.
[21] Beyer M, Aumann J. Effects of Fusarium infection on the aminoacid composition of winter wheat grain. Food Chem2008;111:750–4.
[22] Nicolaisen M, Supronienė S, Nielsen LK, Lazzaro I, Spliid NH,Justesen AF. Real-time PCR for quantification of elevenindividual Fusarium species in cereals. J Microbiol Methods2009;76:234–40.
[23] Fredlund E, Gidlund A, Olsen M, Börjesson T, Spliid NH,Simonsson M. Method evaluation of Fusarium DNA extractionfrom mycelia and wheat for down-stream real-time PCRquantification and correlation tomycotoxin levels. J MicrobialMethods 2008;73:33–40.
[24] Finnie C, Melchior S, Roepstorff P, Svensson B. Proteomeanalysis of grain filling and seed maturation in barley. PlantPhysiol 2002;129:1308–19.
[25] Finnie C, Svensson B. Proteolysis during the isoelectricfocusing step of two-dimensional gel electrophoresis may bea common problem. Anal Biochem 2002;311:182–6.
[26] PopovN, SchmittM, Schulzeck S,MatthiesH. Eine störungsfreieMikromethode zur bestimmung des Proteingehaltes inGewebehomogenaten. Acta Biol Med Ger 1975;34:1441–6.
[27] Jacobsen S, Nesic L, Petersen M, Søndergaard I. Classificationof wheat varieties: use of two-dimensional gel electrophoresisfor varieties that cannot be classified by matrix assistedlaser desorption/ionization-time of flight-mass spectrometryand an artificial neural network. Electrophoresis2001;22:1242–5.
[28] BradfordMM.Arapidandsensitivemethod for thequantificationof microgram quantities of protein utilizing the principle ofprotein-dye binding. Anal Biochem 1976;72:248–54.
[29] Rabilloud T, Charmont S. Detection of proteins ontwo-dimensional electrophoresis gels. Two-dimensional gelelectrophoresis and identification methods. In: Rabilloud T,editor. Proteome research. Berlin, Heidelberg: Springer Verlag;2000. p. 109–10.
[30] Zhang X, Shi L, Shu S, Wang Y, Zhao K, Xu N, et al. Animproved method of sample preparation on AnchorChiptargets for MALDI-MS and MS/MS and its application in theliver proteome project. Proteomics 2007;7:340–9.
[31] Finnie C, Steenholdt T, Roda Noguera O, Knudsen S, Larsen J,Brinch-Pedersen H. Environmental and transgene expressioneffects on the barley seed proteome. Phytochemistry2004;65:1619–27.
[32] Østergaard O, Finnie C, Laugesen S, Roepstorff P, Svensson B.Proteome analysis of barley seeds: identification of majorproteins from two-dimensional gels (pl 4–7). Proteomics2004;4:2437–47.
[33] Cloud GL, Rupe JC. Influence of nitrogen, plant growth stage,and environment on charcoal rot of grain sorghum caused byMacrophomina phaseolina (Tassi) Goid. Plant Soil1994;158:203–10.
[34] Taylor RD, Saparno A, Blackwell B, Anoop V, Gleddie S, TinkerNA, et al. Proteomic analyses of Fusarium graminearum grownunder mycotoxin-inducing conditions. Proteomics2008;8:2256–65.
[35] Miller JD, Blackwell BA. Biosynthesis of 3-acetyldeoxynivalenoland other metabolites by Fusarium culmorum HLK 1503 in astirred jar fermentor. Can J Bot 1986;64:1–5.
[36] Dodd JL. The role of plant stresses in development of cornstalk rots. Plant Dis 1980;64:533–7.
[37] Jackowiak H, Packa D, Wiwart M, Perkowski J. Scanningelectron microscopy of Fusarium damaged kernels of springwheat. Int J Food Microbiol 2005;98:113–23.
752 J O U R N A L O F P R O T E O M I C S 7 3 ( 2 0 1 0 ) 7 4 3 – 7 5 2
[38] Bechtel DB, Kaleikau LA, Gaines RL, Seitz LM. The effects ofFusarium graminearum infection on wheat kernels. CerealChem 1985;62:191–7.
[39] Nightingale MJ, Marchylo BA, Clear RM, Dexter JE, Preston KR.Fusarium Head Blight: effect of fungal proteases on wheatstorage proteins. Cereal Chem 1999;76:150–8.
[40] Schwarz PB, Jones BL, Steffenson BJ. Enzymes associated withFusarium infection of barley. J Am Soc Brew Chem2002;60:130–4.
[41] Pekkarinen AI, Sarlin TH, Laitila AT, Haikara AI, Jones BL.Fusarium species synthesize alkaline proteinases in infestedbarley. J Cereal Sci 2003;37:349–56.
[42] Mehta A, Brasileiro AC, Souza DS, Romano E, Campos MA,Grossi-de-Sá MF, et al. Plant–pathogen interactions: what isproteomics telling us? J FEBS 2008;275:3731–46.
[43] Armstrong PB. Proteases and protease inhibitors: a balance ofactivities in host–pathogen interaction. Immunobiol2006;211:263–81.
[44] Pekkarinen AI, Jones BL. Purification and identification ofbarley (Hordeum vulgare L.) proteins that inhibit the alkalineserine proteinases of Fusarium culmorum. J Agric Food Chem2003;51:1710–7.
[45] Pekkarinen AI, Longstaff C, Jones BL. Kinetics of the inhibitionof Fusarium serine proteinases by barley (Hordeum vulgare L.)inhibitors. J Agric Food Chem 2007;55:2736–42.
[46] Shewry PR, Tatham AS, Halford NG. Nutritional control ofstorage protein synthesis in developing grain of wheat andbarley. Plant Growth Regul 2001;34:105–11.
[47] Ríos G, Pinson-Gadais L, Abecassis J, Zakhia-Rozis N,Lullien-Pellerin V. Assessment of dehulling efficiency toreduce deoxynivalenol and Fusarium level in durum wheatgrains. J Cereal Sci 2009;49:387–9.
[48] Link T, Lohaus G, Heiser I, Mendgen K, Hahn M, Voegele RT.Characterization of a novel NADP+-dependent D-arabitoldehydrogenase from the plant pathogen Uromyces fabae.Biochem J 2005;389:289–95.
[49] Thewes S, Prado-Cabrero A, Prado MM, Tudzynski B, Avalos J.Characterization of a gene in the car cluster of Fusariumfujikuroi that codes for a protein of the carotenoid oxygenasefamily. Mol Gen Genomics 2005;274:217–28.
[50] Tudzynski B. Biosynthesis of gibberellins in Gibberella fujikuroi:biomolecular aspects. Appl Microbiol Biotechnol1999;52:298–310.
Chapter 3
38
Chapter 3
Analysis of early events in the interaction between Fusarium
graminearum and the susceptible barley (Hordeum vulgare) cultivar
Scarlett
RESEARCH ARTICLE
Analysis of early events in the interaction between
Fusarium graminearum and the susceptible barley
(Hordeum vulgare) cultivar Scarlett
Fen Yang1, Jens D. Jensen2, Birte Svensson1, Hans J. L. Jørgensen2, David B. Collinge2
and Christine Finnie1
1 Enzyme and Protein Chemistry, Department of Systems Biology, Technical University of Denmark, Denmark2 Department of Plant Biology and Biotechnology, Faculty of Life Sciences, University of Copenhagen, Denmark
Received: April 14, 2010
Revised: July 7, 2010
Accepted: August 3, 2010
A proteomic analysis was conducted to map the events during the initial stages of the
interaction between the fungal pathogen Fusarium graminearum and the susceptible barley
cultivar Scarlett. Quantification of fungal DNA demonstrated a sharp increase in fungal
biomass in barley spikelets at 3 days after inoculation. This coincided with the appearance of
discrete F. graminearum-induced proteolytic fragments of b-amylase. Based on these results,
analysis of grain proteome changes prior to extensive proteolysis enabled identification of
barley proteins responding early to infection by the fungus. In total, the intensity of 51
protein spots was significantly changed in F. graminearum-infected spikelets and all but one
were identified. These included pathogenesis-related proteins, proteins involved in energy
metabolism, secondary metabolism and protein synthesis. A single fungal protein of
unknown function was identified. Quantitative real-time RT-PCR analysis of selected genes
showed a correlation between high gene expression and detection of the corresponding
proteins. Fungal genes encoding alkaline protease and endothiapepsin were expressed during
1–3 days after inoculation, making them candidates for generation of the observed b-amylase
fragments. These fragments have potential to be developed as proteome-level markers for
fungal infection that are also informative about grain protein quality.
cycle, increased in response to F. graminearum. S-adenosyl-L-
methionine, the major methyl-group donor in the cycle, is
required for the biosynthesis of various phenylpropanoid
derivatives and is also an intermediate in the biosynthesis of
ethylene, both of which are associated with plant defence [2,
27]. The gene encoding S-adenosyl-L-homocysteine hydro-
lase was activated rapidly in cultured cells and leaves of
parsley after treatment with fungal elicitors [27, 28]. An
expansin-like protein was also upregulated, possibly result-
ing in cell wall alterations affecting susceptibility to patho-
gens [29]. Two spots with decreased intensity in
F. graminearum-infected spikelets contained thiamine
biosynthetic enzyme and progesterone 5-b-reductase,
involved in biosynthesis of secondary metabolites thiamine
and cardenolide, respectively, which may affect plant
responses to pathogens [30, 31]. The decrease in abundance
of cell division control protein in two spots in infected plants
suggested that fungal attack may reduce cell division and
thus interfere with grain development.
The proteins identified here are probably among the
most abundant proteins responding to the early stages of
F. graminearum infection. A deeper insight into less abun-
dant proteins would be achieved by applying enrichment
strategies combined with LC-MS/MS for protein identifica-
tion. However, it is worth noting that the identification and
analysis of proteolytic fragments necessitates a gel-based
approach since information about protein mass is lost in
gel-free approaches. In fact, in cases such as F. graminearuminfection, where one of the biological effects is significant
proteolysis, misleading quantification of proteins may result
from gel-free analysis.
In conclusion, by characterising the degree of F. grami-nearum colonisation based on measurements of fungal
biomass and fungal-induced proteolysis it was possible to
carry out proteome analysis at a well-defined stage prior to
extensive degradation of plant proteins. This enabled iden-
tification of barley proteins responding early to infection by
the fungus. These changes were compared with the induc-
tion of PR-gene expression in the first days of infection. The
results suggest that initially there is an increased energy
metabolism in infected seeds that may aid the growth of the
fungus. Protease genes expressed by the fungus are likely to
cause the massive degradation of plant proteins previously
observed. Proteolytic activity in infected grains causes the
early appearance of discrete fragments of b-amylase, which
can serve as a proteome-level indicator of infection and grain
protein quality.
We thank Evan Evans (University of Tasmania, Australia)for providing antibodies, Corby Kistler (University of Minnesota,US) for providing F. graminearum PH1 spores and Cb GowdaRayapuram (University of Copenhagen, Denmark) for design-ing and providing primers for PR-1b, peroxidase, oxalate oxidaseand b-1,3-glucanase 2a. This work is funded by the Directoratefor Food, Fisheries and Agri Business (DFFE) grant ‘‘Fusarium
disease resistance–toxins and feed quality,’’ Plant BiotechDenmark, the Centre for Advanced Food Studies (LMC) and aPhD stipend from the Technical University of Denmark.
The authors have declared no conflict of interest.
5 References
[1] Walter, S., Nicholson, P., Doohan, F. M., Action and reaction
of host and pathogen during Fusarium head blight disease.
New Phytol. 2009, 185, 54–66.
[2] Boddu, J., Cho, S., Kruger, W. M., Muehlbauer, G. J.,
Transcriptome analysis of the barley-Fusarium gram-
a Spot numbers a, b and e refer to Fig. 2 (albumin fractions), spot numbers c, d and f refer to Fig. 3 (gliadin fractions).
b +: spots increasing in intensity under stress treatments; -: spots decreasing in intensity under stress treatments.
22
c Significant MS score is above 71 for NCBI and HvGI.
d gi: accession number in NCBI; TC, BQ, BE, BJ: accession number in the TaGI wheat gene index Release 11.0.
e When the identification was based on EST sequence, the organism with the most homologous sequence is given. The theoretical pI and MW are calculated from the
homologous sequence of that organism. f Significant MS/MS score is above 26 for TaGI.
Appendix II
116
Appendix II
Secretome analysis of YAP- or AP-knockout mutant of Fusarium
graminearum
Collaboration with Ph.D. student Jens Due Jensen, Department of Plant Biology and Biotechnology, University
of Copenhagen
YAP and AP are known as transcription factors which regulate genes encoding antioxidants such as catalases and
superoxide dismutases and are required for the response to oxidative stress (personal communication with Jens
Due Jensen). Two Fusarium graminearum mutants have been knocked out YAP and AP genes, respectively,
constructed by Jens Due Jensen, to investigate the role of antioxidants in pathogenesis. In in planta experiments,
these two mutants are more aggressive than wildtype strain during infection of wheat. In order to gain some
insights into the molecular mechanism of infection in the mutants, we conducted in vitro secretome analysis with
the substrate of wheat grain flour due to difficulty of obtaining fungal proteins in planta and the important role
of secreted proteins in pathogenicity. The methods of isolation of fungal secreted proteins, 2-DE, image analysis
and protein identification were performed as described in Chapter 4. In total, 38 and 34 spots changed in
intensity in AP and YAP mutants compared to the wildtype, of which 35 and 31 spots were identified,
respectively. There was one spot identified as wheat abundant protein serpin in either mutant. The identified
fungal proteins were mainly involved in the degradation of plant substrate cell wall, starch and protein. The 2-D
patterns of mutants as well as wildtype and protein identification were shown in Fig. S1 and Table S1,
respectively.
Appendix II
117
Fig. S1. 2-DE Sypro-Ruby stained gels from the secretome of wildtype (PH1), AP-knockout and YAP-knockout F. graminearum grown in
medium containing wheat grain flour. Molecular size markers and pI ranges are indicated. Protein spots changing in intensity between the
AP or YAP mutant and wildtype are numbered as Axx and Yxx, respectively.
Appendix II
118
Table S1. Identification list of F. graminearum proteins varying between mutants and wildtype
Y635d -1.5 85 13 27 6.64/6.01 50.57/34.90 TC253068 Serpin No
Appendix II
120
a +: increasing in intensity in the mutants; -: decreasing in intensity in the mutants b FGSG: accession number in Fusarium graminearum gene index from Broad Institute; TC: accession number in the TaGI wheat gene index Release 11.0 c If the signal peptide is contained, the theoretical pI and MW are calculated after removal of signal peptide d At least one peptide additionally confirmed by MS/MS analysis e Identified by MS/MS analysis